Enterococci are the third most common cause of endocarditis and have been known to cause this infection for over 100 years. While, in the past, infections due to Enterococcus faecium were infrequent, this organism has increased markedly in frequency and in importance in the hospital setting where it is now one of the most difficult to treat organisms. The recent generation of the genome sequence of our E. faecium endocarditis isolate TX16 (=DO), part of a collaboration between the PI, G. Weinstock and the JGI, allowed us to identify Acm, the primary adhesin to collagen of E. faecium. Although acm is present in all [unreadable] faecium isolates, collagen adherence, encoded by acm, is displayed by only 3/60 (5%) non-clinical isolates vs. 36/63 (57%) clinical isolates (p < 0.0001). As a partial explanation for these phenotypic differences, we have shown that acm is frequently present as a pseudogene in fecal isolates from healthy humans or animals; we have a\soshown that acm is intact in all endocarditis isolates studied to date, and there is evidence of Acm expression during infections in man, even when not expressed in vitro. Recent preliminary data found that a non-adhering acm deletion mutant that we generated from a human [unreadable] faecium endocarditis isolate with a new ts vector we constructed for this project, is attenuated in its ability to cause endocarditis in rats. We have also identified 14 other potential extracellular matrix (ECM) adhesins including some with structural features characteristic of cell-wall anchored proteins plus motifs (Ig-like folds) seen in other ECM binding proteins, three of which are in loci with sortase-like genes. Recombinant protein extracts of two of these adhere to fibrinqgen or fibronectin (one each), consistent with decreased fibronectin binding seen with a disruption mutation in one. The long term goals of our work are to translate in vitro observations relating to these genes and ECM adherence into knowledge useful for combating enterococcal infections, among which, endocarditis is the most problematic. Specific Aims related to Acm are to confirm its importance in animal models; to characterize the binding region(s) of Acm and the mechanism for its differences in binding vs.the collagen adhesin Ace of [unreadable] faecalis; to investigate genetic and/or environmental conditions that influence Acm production; to determine if anti- Acm antibodies can prevent collagen adherence and/or detach bacteria bound to collagen; and to explore passive and active immunization for prevention of [unreadable] faecium endocarditis and other infections in animals. The aims relating to the other (potential) adhesins of [unreadable] faecium are to use a multi-faceted approach to understand the causes and relevance of [unreadable] faecium adherence to other ECM proteins. These aims will be addressed by introducing mutations into the (putative) adhesin genes of adherence-positive strain(s), by cloning these genes into non-adherent host(s), by testing recombinant proteins for in vitro adherence and with infected patients' sera for evidence of in vivo expression of these proteins and, time permitting, initiating experiments that would address the role of these adhesins in vivo. The broad hypotheses of this proposal are (1) that (at least) some of these confirmed and putative adhesins contribute to making [unreadable] faecium a successful cause of endocarditis and other infections, (2) that some are responsible for the "untaming" of nosocomial [unreadable] faecium isolates seen in recent years, and (3) that a better understanding of these confirmed and putative adhesins will help lead to modalities to prevent and/or reverse their effects.